Why don't you factor tilt?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Random Walker
Posts: 5312
Joined: Fri Feb 23, 2007 8:21 pm

Re: Why don't you factor tilt?

Post by Random Walker »

km91 wrote: Thu Sep 22, 2022 3:17 pm
abc132 wrote: Thu Sep 22, 2022 2:38 pm I was thinking about the independence of factors and wondering why there would be any products suggested that were combinations of factors.

For example, X% (small minus big) and Y% (value minus growth) should be a much better way to invest than SCV. The former makes it easy to maintain factor weightings and the latter makes it impossible.

The only thing I can come up with is we may say factors are independent but we invest as if they are not independent - chasing the combination that worked well recently.

The profitability + small post in this thread is a good example of what I am questioning, along with lots of comments (in general in the last few years) that you just need the right combination of factors. All of these seem to speak of factors not actually being independent.

Which is it?
The FF factor model describes stock returns as a linear combination (weighted average) of the factor loadings and return premiums. Each factor independently describes some dimension of stock returns and as a whole any given portfolio can be described as a combination of the factors. The larger the tilt away from the MKT premium and towards the other factors, the more diversified the portfolio should become.

I think it would be practically impossible to create a portfolio with a loading of 100% to value and 0% to any other factor, even in a long short construction. The factor model tells us returns are driven by a combination of the factors. We are not concerned about isolating any single factor, but rather creating a portfolio that has less of its return driven by MKT and more of the return driven by alternative factors
Agreed. Also don’t want to have multiple single factor funds where one fund might be buying a stock and another fund selling the same stock. For example a momentum fund might be selling a stock that is going down. A value fund might want to purchase the same stock as it enters its buy range. Better to avoid trading costs and put the factors into one fund. Better to have a single fund that owns stocks ranking B in both categories than single factor funds that will own A stocks for one category that get D grade in the other.

Dave
abc132
Posts: 1122
Joined: Thu Oct 18, 2018 1:11 am

Re: Why don't you factor tilt?

Post by abc132 »

Random Walker wrote: Thu Sep 22, 2022 5:15 pm
km91 wrote: Thu Sep 22, 2022 3:17 pm
abc132 wrote: Thu Sep 22, 2022 2:38 pm I was thinking about the independence of factors and wondering why there would be any products suggested that were combinations of factors.

For example, X% (small minus big) and Y% (value minus growth) should be a much better way to invest than SCV. The former makes it easy to maintain factor weightings and the latter makes it impossible.

The only thing I can come up with is we may say factors are independent but we invest as if they are not independent - chasing the combination that worked well recently.

The profitability + small post in this thread is a good example of what I am questioning, along with lots of comments (in general in the last few years) that you just need the right combination of factors. All of these seem to speak of factors not actually being independent.

Which is it?
The FF factor model describes stock returns as a linear combination (weighted average) of the factor loadings and return premiums. Each factor independently describes some dimension of stock returns and as a whole any given portfolio can be described as a combination of the factors. The larger the tilt away from the MKT premium and towards the other factors, the more diversified the portfolio should become.

I think it would be practically impossible to create a portfolio with a loading of 100% to value and 0% to any other factor, even in a long short construction. The factor model tells us returns are driven by a combination of the factors. We are not concerned about isolating any single factor, but rather creating a portfolio that has less of its return driven by MKT and more of the return driven by alternative factors
Agreed. Also don’t want to have multiple single factor funds where one fund might be buying a stock and another fund selling the same stock. For example a momentum fund might be selling a stock that is going down. A value fund might want to purchase the same stock as it enters its buy range. Better to avoid trading costs and put the factors into one fund. Better to have a single fund that owns stocks ranking B in both categories than single factor funds that will own A stocks for one category that get D grade in the other.

Dave
If only someone patented a way to not have to do that...
Apathizer
Posts: 1326
Joined: Sun Sep 26, 2021 2:56 pm

Re: Why don't you factor tilt?

Post by Apathizer »

abc132 wrote: Thu Sep 22, 2022 5:00 pm
Apathizer wrote: Thu Sep 22, 2022 4:48 pm
abc132 wrote: Thu Sep 22, 2022 4:06 pmYes, I want to benefit from the diversification of which factor over or underperforms if I believe they have an expected value. Take all the SCV people that think expected returns are higher now than ever. You are stuck with all 4 factors regardless of what has done well recently, and you have less SCV in this example when expected returns may be higher.
We don't know which and when certain factors will out-perform others. All we can do is guesstimate the likelihood based on our understanding of asset pricing and historical performance. That's why diversifying across the 5 most relevant makes sense.

It's basic probability. As I've said, 5 independent potential return sources are less likely to perform poorly than only 1. To use a card analogy you're more likely to win if you play 5 random hands than you are if you only play 1.
abc132 wrote: Thu Sep 22, 2022 4:31 pmI would argue factor investors today have something like 2.4 sources of risk with lower expected returns, all of which gives zero expected advantage.
What are you basing this on? This statement is counterfactual. It's like saying a non-smoker is more likely to get cancer than a smoker. That doesn't mean non-smokers will get cancer, but they're less likely to than smokers. Probabilistically we know this is true.
Roll 5 dice and they are more likely to be nearer the average than rolling one die.
Roll an infinite number of dice and you should approach the average, or expected value.

You are saying you can take the average of 5 dice and get the same diversification of rolling them independently.

Having an infinite number of independent factors would merely approach the expected value, and this would perform exactly as a bond but with separate risk (the sum of expected values changes, and your prior purchase value adjusts accordingly).

n-factors can't provide more diversification than including bonds since the infinite number of factors approaches the bond case. At best, we would get an additional bond-like asset. We should expect factors to act as something between a bond and a stock with lower expected performance from the extra diversification.
No, because not all factors are equally relevant. While there are a multitude of factors, not all are statistically significant, but some are. The best available information shows value, size, profitability, and reinvestment seem to be the most statistically significant. It could be entirely coincidental, but that's unlikely. Considering all this, slanting towards these factors seems likely to improve returns, though it's not guaranteed.
ROTH: 40% AVUS, 30% DFAX, 30% BNDW. Taxable: 50% BNDW, 30% AVUS, 20% DFAX.
abc132
Posts: 1122
Joined: Thu Oct 18, 2018 1:11 am

Re: Why don't you factor tilt?

Post by abc132 »

Apathizer wrote: Thu Sep 22, 2022 5:19 pm No, because not all factors are equally relevant. While there are a multitude of factors, not all are statistically significant, but some are. The best available information shows value, size, profitability, and reinvestment seem to be the most statistically significant. It could be entirely coincidental, but that's unlikely. Considering all this, slanting towards these factors seems likely to improve returns, though it's not guaranteed.
I am playing the "what-if" of what happens with an infinite number of factors to show that any number of factors provide less diversification than an equivalent amount of bonds. You may be able to juice risk vs reward, but that depends to what extent the best of the past continue to be the best of the future. If you consider mean and deviation and that we are choosing the best of the past, the best of the past should not be expected to give the same returns in the future.

As a simple example we could be loading up on large minus small with changes in our history for how large and small companies were treated.
km91
Posts: 279
Joined: Wed Oct 13, 2021 12:32 pm

Re: Why don't you factor tilt?

Post by km91 »

abc132 wrote: Thu Sep 22, 2022 5:18 pm If only someone patented a way to not have to do that...
Lol. I don't think the take away should be that cap weighted TSM is insufficient. I think the take away from the factor models should be that for those investors who have a willingness to take on the risk and who believe the factors represent persistent premiums that won't be competed or arbitraged out of the market, that there are additional degrees of equity diversification beyond just cap weight
abc132
Posts: 1122
Joined: Thu Oct 18, 2018 1:11 am

Re: Why don't you factor tilt?

Post by abc132 »

km91 wrote: Thu Sep 22, 2022 5:47 pm
abc132 wrote: Thu Sep 22, 2022 5:18 pm If only someone patented a way to not have to do that...
Lol. I don't think the take away should be that cap weighted TSM is insufficient. I think the take away from the factor models should be that for those investors who have a willingness to take on the risk and who believe the factors represent persistent premiums that won't be competed or arbitraged out of the market, that there are additional degrees of equity diversification beyond just cap weight
Absolutely, but I do not believe most are investing for these reasons, at least not in writing. Things like 5 factors vs 1 factor are the canary in the coal mine for investor behavior - that they believe they are taking on less expected risk and getting more expected reward. Factors + additional bonds might do the trick, but it is the bonds that provide less risk not the 5 independent factors.
km91
Posts: 279
Joined: Wed Oct 13, 2021 12:32 pm

Re: Why don't you factor tilt?

Post by km91 »

abc132 wrote: Thu Sep 22, 2022 5:55 pm
km91 wrote: Thu Sep 22, 2022 5:47 pm
abc132 wrote: Thu Sep 22, 2022 5:18 pm If only someone patented a way to not have to do that...
Lol. I don't think the take away should be that cap weighted TSM is insufficient. I think the take away from the factor models should be that for those investors who have a willingness to take on the risk and who believe the factors represent persistent premiums that won't be competed or arbitraged out of the market, that there are additional degrees of equity diversification beyond just cap weight
Absolutely, but I do not believe most are investing for these reasons, at least not in writing. Things like 5 factors vs 1 factor are the canary in the coal mine for investor behavior - that they believe they are taking on less expected risk and getting more expected reward. Factors + additional bonds might do the trick, but it is the bonds that provide less risk not the 5 independent factors.

I think framing it in comparison to bonds is the wrong way to look at it. Bonds are diversification across asset classes, factors are diversification within an asset class. I think the better way is to think of a 100% equity portfolio. At the end of the day, no matter how much factor tilt you have in the portfolio you are still exposed to "equity" risk as a whole, however you choose to define it. The insight is that you can improve the risk/return profile of the equity side of the portfolio and achieve a more robust equity return stream by diversifying into additional factors beyond cap weighted total stock market. I find it hard to believe that investors as a whole think they are getting more return with less risk by tilting towards the factors. If that was true SCV would've performed much better than it has over the last 10 years as investors piled in to get their supposed free lunch
Apathizer
Posts: 1326
Joined: Sun Sep 26, 2021 2:56 pm

Re: Why don't you factor tilt?

Post by Apathizer »

abc132 wrote: Thu Sep 22, 2022 5:39 pm
Apathizer wrote: Thu Sep 22, 2022 5:19 pm No, because not all factors are equally relevant. While there are a multitude of factors, not all are statistically significant, but some are. The best available information shows value, size, profitability, and reinvestment seem to be the most statistically significant. It could be entirely coincidental, but that's unlikely. Considering all this, slanting towards these factors seems likely to improve returns, though it's not guaranteed.
I am playing the "what-if" of what happens with an infinite number of factors to show that any number of factors provide less diversification than an equivalent amount of bonds. You may be able to juice risk vs reward, but that depends to what extent the best of the past continue to be the best of the future. If you consider mean and deviation and that we are choosing the best of the past, the best of the past should not be expected to give the same returns in the future.

As a simple example we could be loading up on large minus small with changes in our history for how large and small companies were treated.
I'm not interested in entirely conjectural 'what-ifs'. I'm in data-based probabilities. A factor-slanted equity and bonds is more well-diversified than a cap weight portfolio with the equivalent bond allocation since the former provides 6 independent potential return sources. To illustrate this we can compare a 50% factor-slanted equities/50% bonds portfolio with a 50% cap weight equities/50% bonds portfolio 2000-2009.
https://www.portfoliovisualizer.com/bac ... tion6_3=40

The factor-slanted portfolio annual percentage is about 1% better, which is fairly significant over a decade. Of course we didn't know this would happen and can't know what happens in the future, but even with 50% equities/50% bonds portfolio factor-slants can be provide beneficial diversification.
km91 wrote: Thu Sep 22, 2022 6:19 pm
abc132 wrote: Thu Sep 22, 2022 5:55 pm
km91 wrote: Thu Sep 22, 2022 5:47 pm
abc132 wrote: Thu Sep 22, 2022 5:18 pm If only someone patented a way to not have to do that...
Lol. I don't think the take away should be that cap weighted TSM is insufficient. I think the take away from the factor models should be that for those investors who have a willingness to take on the risk and who believe the factors represent persistent premiums that won't be competed or arbitraged out of the market, that there are additional degrees of equity diversification beyond just cap weight
Absolutely, but I do not believe most are investing for these reasons, at least not in writing. Things like 5 factors vs 1 factor are the canary in the coal mine for investor behavior - that they believe they are taking on less expected risk and getting more expected reward. Factors + additional bonds might do the trick, but it is the bonds that provide less risk not the 5 independent factors.

I think framing it in comparison to bonds is the wrong way to look at it. Bonds are diversification across asset classes, factors are diversification within an asset class. I think the better way is to think of a 100% equity portfolio. At the end of the day, no matter how much factor tilt you have in the portfolio you are still exposed to "equity" risk as a whole, however you choose to define it. The insight is that you can improve the risk/return profile of the equity side of the portfolio and achieve a more robust equity return stream by diversifying into additional factors beyond cap weighted total stock market. I find it hard to believe that investors as a whole think they are getting more return with less risk by tilting towards the factors. If that was true SCV would've performed much better than it has over the last 10 years as investors piled in to get their supposed free lunch
Exactly. Equities and bonds serve different purposes. Equities are for maximizing long-term returns while bonds are intended to provide fairly consistent short-term returns (though that hasn't worked out the last couple years).
ROTH: 40% AVUS, 30% DFAX, 30% BNDW. Taxable: 50% BNDW, 30% AVUS, 20% DFAX.
abc132
Posts: 1122
Joined: Thu Oct 18, 2018 1:11 am

Re: Why don't you factor tilt?

Post by abc132 »

Apathizer wrote: Thu Sep 22, 2022 6:22 pm
abc132 wrote: Thu Sep 22, 2022 5:39 pm
Apathizer wrote: Thu Sep 22, 2022 5:19 pm No, because not all factors are equally relevant. While there are a multitude of factors, not all are statistically significant, but some are. The best available information shows value, size, profitability, and reinvestment seem to be the most statistically significant. It could be entirely coincidental, but that's unlikely. Considering all this, slanting towards these factors seems likely to improve returns, though it's not guaranteed.
I am playing the "what-if" of what happens with an infinite number of factors to show that any number of factors provide less diversification than an equivalent amount of bonds. You may be able to juice risk vs reward, but that depends to what extent the best of the past continue to be the best of the future. If you consider mean and deviation and that we are choosing the best of the past, the best of the past should not be expected to give the same returns in the future.

As a simple example we could be loading up on large minus small with changes in our history for how large and small companies were treated.
I'm not interested in entirely conjectural 'what-ifs'. I'm in data-based probabilities.
You can simply double your number of factors in your data, copy the performance of another factor, and look at what happens. Or not if you don't what to understand why 6 should not be compared to 1.

Apathizer wrote: Thu Sep 22, 2022 6:22 pm A factor-slanted equity and bonds is more well-diversified than a cap weight portfolio with the equivalent bond allocation since the former provides 6 independent potential return sources. To illustrate this we can compare a 50% factor-slanted equities/50% bonds portfolio with a 50% cap weight equities/50% bonds portfolio 2000-2009.
https://www.portfoliovisualizer.com/bac ... tion6_3=40

The factor-slanted portfolio annual percentage is about 1% better, which is fairly significant over a decade. Of course we didn't know this would happen and can't know what happens in the future, but even with 50% equities/50% bonds portfolio factor-slants can be provide beneficial diversification.
Why didn't you throw in the large minus small or growth minus value factor? You data mined what worked well and found out that for any data set we can always find a subset, concentrate it, and beat the average. That is a required attribute of any data set, and as a bonus you can always declare diversification (X factors vs 1). This would be true for completely random data and proves nothing.
User avatar
Taylor Larimore
Advisory Board
Posts: 31577
Joined: Tue Feb 27, 2007 8:09 pm
Location: Miami FL

Re: Why don't you factor tilt?

Post by Taylor Larimore »

Bogleheads:

For what it's worth, I looked-up the longest Morningstar return (3 years) for Vanguard U.S. Multifactor ETF (VMMF). It was 8.49%.

Meanwhile the 3 year return for Vanguard U.S. Total Market Index ETF (VTI) was 9.34%. Another reason I'm happy I don't "factor tilt."

What Experts Say About Total Market Index Funds

Best wishes.
Taylor
Jack Bogle's Words of Wisdom: "One of the seemingly indestructible myths of investing is that stocks with small market capitalizations outpace stocks with large market capitalizations over time."
"Simplicity is the master key to financial success." -- Jack Bogle
abc132
Posts: 1122
Joined: Thu Oct 18, 2018 1:11 am

Re: Why don't you factor tilt?

Post by abc132 »

Taylor Larimore wrote: Thu Sep 22, 2022 7:21 pm Bogleheads:

For what it's worth, I looked-up the longest Morningstar return (3 years) for Vanguard U.S. Multifactor ETF (VMMF). It was 8.49%.

Meanwhile the 3 year return for Vanguard U.S. Total Market Index ETF (VTI) was 9.34%. Another reason I'm happy I don't "factor tilt."

What Experts Say About Total Market Index Funds

Best wishes.
Taylor
Jack Bogle's Words of Wisdom: "One of the seemingly indestructible myths of investing is that stocks with small market capitalizations outpace stocks with large market capitalizations over time."
I keep looking back at Aug 2018 when I first diversified out of 100% stocks (US, Int'l, Emerging), and so far none of the recommended portfolios have done better that stocks + bonds. TIPS vs nominals, factor or not, nothing has really mattered much. I have just one timeline so anything could happen from here. I look forward to really simplifying once my bonds get me all the way to social security, or 59.5 when I can move/combine everything.
Apathizer
Posts: 1326
Joined: Sun Sep 26, 2021 2:56 pm

Re: Why don't you factor tilt?

Post by Apathizer »

abc132 wrote: Thu Sep 22, 2022 6:33 pm Why didn't you throw in the large minus small or growth minus value factor? You data mined what worked well and found out that for any data set we can always find a subset, concentrate it, and beat the average. That is a required attribute of any data set, and as a bonus you can always declare diversification (X factors vs 1). This would be true for completely random data and proves nothing.
That was the closet factor-diversified portfolio I could do on portfolio visualizer. It's a reasonable approximation of a factor slanted portfolio.
ROTH: 40% AVUS, 30% DFAX, 30% BNDW. Taxable: 50% BNDW, 30% AVUS, 20% DFAX.
Apathizer
Posts: 1326
Joined: Sun Sep 26, 2021 2:56 pm

Re: Why don't you factor tilt?

Post by Apathizer »

Taylor Larimore wrote: Thu Sep 22, 2022 7:21 pm Bogleheads:

For what it's worth, I looked-up the longest Morningstar return (3 years) for Vanguard U.S. Multifactor ETF (VMMF). It was 8.49%.

Meanwhile the 3 year return for Vanguard U.S. Total Market Index ETF (VTI) was 9.34%. Another reason I'm happy I don't "factor tilt."

What Experts Say About Total Market Index Funds

Best wishes.
Taylor
Jack Bogle's Words of Wisdom: "One of the seemingly indestructible myths of investing is that stocks with small market capitalizations outpace stocks with large market capitalizations over time."
While Vanguard is great for total market index funds, they aren't very astute with factors. As experts opinion, that's the lowest standard of evidence. There are also plenty of other experts who make compelling arguments for a factor-slanted portfolio.
ROTH: 40% AVUS, 30% DFAX, 30% BNDW. Taxable: 50% BNDW, 30% AVUS, 20% DFAX.
abc132
Posts: 1122
Joined: Thu Oct 18, 2018 1:11 am

Re: Why don't you factor tilt?

Post by abc132 »

Apathizer wrote: Thu Sep 22, 2022 7:44 pm
abc132 wrote: Thu Sep 22, 2022 6:33 pm Why didn't you throw in the large minus small or growth minus value factor? You data mined what worked well and found out that for any data set we can always find a subset, concentrate it, and beat the average. That is a required attribute of any data set, and as a bonus you can always declare diversification (X factors vs 1). This would be true for completely random data and proves nothing.
That was the closet factor-diversified portfolio I could do on portfolio visualizer. It's a reasonable approximation of a factor slanted portfolio.
Yes, and as I explained it is a requirement for any data set to be able to both
- concentrate something to do better than the average
- declare that portfolio more diversified

It is impossible not to be able to beat the market after the fact.

Why don't you try out some large minus small (size factor) and growth minus value (growth factor) portfolios and see how they did historically. Why are you picking just the factor choices that back tested well?

Or the same portfolios 2006-2022 which paints the opposite picture?

Remember that we can't use results to declare the choice good or bad, as we never know
- was the expected value wrong?
- was it just poor luck?
Last edited by abc132 on Thu Sep 22, 2022 8:32 pm, edited 1 time in total.
User avatar
Beensabu
Posts: 2602
Joined: Sun Aug 14, 2016 3:22 pm

Re: Why don't you factor tilt?

Post by Beensabu »

homebuyer6426 wrote: Wed Sep 21, 2022 8:53 am I don't factor tilt because I sector tilt. After looking at more than 50 years of sector average returns and volatility data. Time will tell if it pays off, but if it doesn't, I'll still be okay.
Hello, fellow poster from page 7!

I also do this. I even found a two-in-one to do it with! By complete accident. But in retrospect, it's a two-fer. It's apparently entirely possible to make an okay and ex-post rationale decision for an initially arguably dumb reason (like "cheap! so much cheaper than most of the other times!").

And then, because a bunch of people were always saying "well, the factors change or multiply into a plethora of additional factors depending on the data in the time period being looked at" (or things that I pretty much interpreted as that), I thought "huh..." Could it be that ~70% of excess return over the "risk-free" investment is explained by market beta and the remaining ~30% by other things that are different depending on the time period being looked at? Is it possible? And then I went, "sure, why not?" Just like that. And then I started investing in the anti-potential-past-factor because I figured it would have its day at some point. And since it hadn't yet, for a really long time, I thought the chances of it having its day during my future timeline was probably better than the other things that had been bid up maybe slightly too much because they'd been pointed out (or been doing super good lately). It's not academic at all. It's observational.

You know how sometimes you're like "I wonder..." and you just sit back and watch to see what happens? And other times you're like "I wonder..." and it makes enough sense that spectating simply won't do? It's kind of like that.

Anyway, it's comforting to see someone else admit to sector tilting. It's so... embracing of unsystematic risk as a diversifying element. I'm not throwing shade. I'm doing the same thing. It feels like it would be a super bad thing in another day and age, but with where we are right now, looking at misalignment of fundamentals vs. pricing, it actually kind of makes sense.
"The only thing that makes life possible is permanent, intolerable uncertainty; not knowing what comes next." ~Ursula LeGuin
Apathizer
Posts: 1326
Joined: Sun Sep 26, 2021 2:56 pm

Re: Why don't you factor tilt?

Post by Apathizer »

abc132 wrote: Thu Sep 22, 2022 7:58 pm Why don't you try out some large minus small (size factor) and growth minus value (growth factor) portfolios and see how they did historically. Why are you picking just the factor choices that back tested well?
OK. I'll compare the available long-term performance of the US TSM, large growth, and small value. Small value out-performs them both by about 3%. There's little difference between TSM and large growth since that's the plurality weighting. This is exactly the point. The TSM is dominated by large growth, while a factor-slanted portfolio provides more diversity and more return potential.
https://www.portfoliovisualizer.com/bac ... ion3_3=100
ROTH: 40% AVUS, 30% DFAX, 30% BNDW. Taxable: 50% BNDW, 30% AVUS, 20% DFAX.
abc132
Posts: 1122
Joined: Thu Oct 18, 2018 1:11 am

Re: Why don't you factor tilt?

Post by abc132 »

Apathizer wrote: Thu Sep 22, 2022 9:22 pm
abc132 wrote: Thu Sep 22, 2022 7:58 pm Why don't you try out some large minus small (size factor) and growth minus value (growth factor) portfolios and see how they did historically. Why are you picking just the factor choices that back tested well?
OK. I'll compare the available long-term performance of the US TSM, large growth, and small value. Small value out-performs them both by about 3%. There's little difference between TSM and large growth since that's the plurality weighting. This is exactly the point. The TSM is dominated by large growth, while a factor-slanted portfolio provides more diversity and more return potential.
https://www.portfoliovisualizer.com/bac ... ion3_3=100
The factor I listed should have underperformed by 3% annually. I was looking for that factors (large minus small) and (growth minus value).

You took that thing that was performing well and has not done much since investors moved into it. Early adopters and fund managers benefit, and investors lose to fees. We see SCV losing in the 2006-2022 data on a performance basis, and even more on a risk adjusted basis.

I suspect a lot of Bogleheads will do the same with the US market, expecting that the past must repeat itself and using 50 year back tests when US underperforms for an extended period. How could anyone factor investing complain about US-only portfolios after doing the same thing themselves?
rkhusky
Posts: 13510
Joined: Thu Aug 18, 2011 8:09 pm

Re: Why don't you factor tilt?

Post by rkhusky »

Apathizer wrote: Thu Sep 22, 2022 12:33 pm The 5 factors seem weakly correlated and largely independent, so a incorporating factors could improve reliability since any of them might perform well independently at any given time. That's my point at the beginning of this post.
Only the long-short factors are independent, but no one invests in them. The long-only factors in which people invest are strongly correlated.
rkhusky
Posts: 13510
Joined: Thu Aug 18, 2011 8:09 pm

Re: Why don't you factor tilt?

Post by rkhusky »

km91 wrote: Thu Sep 22, 2022 3:17 pmThe larger the tilt away from the MKT premium and towards the other factors, the more diversified the portfolio should become.
Just the opposite is true. The more long-only factors you apply to your portfolio, the less diversified and more concentrated your portfolio becomes. The risk associated with each factor gets compounded.
homebuyer6426
Posts: 966
Joined: Tue Feb 07, 2017 9:08 am

Re: Why don't you factor tilt?

Post by homebuyer6426 »

Beensabu wrote: Thu Sep 22, 2022 8:14 pm
homebuyer6426 wrote: Wed Sep 21, 2022 8:53 am I don't factor tilt because I sector tilt. After looking at more than 50 years of sector average returns and volatility data. Time will tell if it pays off, but if it doesn't, I'll still be okay.
Hello, fellow poster from page 7!

I also do this. I even found a two-in-one to do it with! By complete accident. But in retrospect, it's a two-fer. It's apparently entirely possible to make an okay and ex-post rationale decision for an initially arguably dumb reason (like "cheap! so much cheaper than most of the other times!").

And then, because a bunch of people were always saying "well, the factors change or multiply into a plethora of additional factors depending on the data in the time period being looked at" (or things that I pretty much interpreted as that), I thought "huh..." Could it be that ~70% of excess return over the "risk-free" investment is explained by market beta and the remaining ~30% by other things that are different depending on the time period being looked at? Is it possible? And then I went, "sure, why not?" Just like that. And then I started investing in the anti-potential-past-factor because I figured it would have its day at some point. And since it hadn't yet, for a really long time, I thought the chances of it having its day during my future timeline was probably better than the other things that had been bid up maybe slightly too much because they'd been pointed out (or been doing super good lately). It's not academic at all. It's observational.

You know how sometimes you're like "I wonder..." and you just sit back and watch to see what happens? And other times you're like "I wonder..." and it makes enough sense that spectating simply won't do? It's kind of like that.

Anyway, it's comforting to see someone else admit to sector tilting. It's so... embracing of unsystematic risk as a diversifying element. I'm not throwing shade. I'm doing the same thing. It feels like it would be a super bad thing in another day and age, but with where we are right now, looking at misalignment of fundamentals vs. pricing, it actually kind of makes sense.
What I can say is that I really value long-term data. I have found sector data and analysis going back to 1960 and earlier. And I was really surprised at what I found because it ran contrary to a lot of the common wisdom of this forum. On the flip side, many people here recommend owning international but it's very hard to find long-term data on it.

I never felt doggedly holding onto a set of principles to the point of shutting out other possibilities was a good idea, even if those principles, on average, are much better than the default reckless behavior. There has to be some room for experimentation and independent analysis. We aren't meant to find a "good enough" set of rules and then shut off our brains. I wish you good luck in achieving your goals.
User avatar
burritoLover
Posts: 2731
Joined: Sun Jul 05, 2020 12:13 pm

Re: Why don't you factor tilt?

Post by burritoLover »

rkhusky wrote: Thu Sep 22, 2022 10:58 pm Only the long-short factors are independent, but no one invests in them. The long-only factors in which people invest are strongly correlated.
rkhusky wrote: Thu Sep 22, 2022 11:09 pm Just the opposite is true. The more long-only factors you apply to your portfolio, the less diversified and more concentrated your portfolio becomes. The risk associated with each factor gets compounded.
https://alphaarchitect.com/2019/11/do-l ... s-deliver/
Image
Random Walker
Posts: 5312
Joined: Fri Feb 23, 2007 8:21 pm

Re: Why don't you factor tilt?

Post by Random Walker »

rkhusky wrote: Thu Sep 22, 2022 10:58 pm
Apathizer wrote: Thu Sep 22, 2022 12:33 pm The 5 factors seem weakly correlated and largely independent, so a incorporating factors could improve reliability since any of them might perform well independently at any given time. That's my point at the beginning of this post.
Only the long-short factors are independent, but no one invests in them. The long-only factors in which people invest are strongly correlated.
The factors are by definition long-short. There are no “long only factors”. But most all of us invest through long only funds that give us access to the factors. Since the factors are definitionally long-short and we invest long only, our long only factor funds typically give us around half (the long side) the factor exposure of the pure factor. The loadings measure the depth of factor exposure for a factor fund. A typical SV fund might have a market exposure of 1.0-1.05, size exposure 0.6-0.8, value exposure 0.4-0.6. Note that a SV fund might have slightly more exposure to market beta than a TSM fund, which by definition is 1.0. That means a SV fund has exposure to the unique and independent risks of size and value and also some increased risk to the same market risk we are trying to diversify away from. This is more reason to factor tilt concomitant with decreasing overall equity allocation. It is because of the market factor exposure of about 1 that SV funds still retain substantial correlation to TSM.

Dave
Random Walker
Posts: 5312
Joined: Fri Feb 23, 2007 8:21 pm

Re: Why don't you factor tilt?

Post by Random Walker »

rkhusky wrote: Thu Sep 22, 2022 11:09 pm
km91 wrote: Thu Sep 22, 2022 3:17 pmThe larger the tilt away from the MKT premium and towards the other factors, the more diversified the portfolio should become.
Just the opposite is true. The more long-only factors you apply to your portfolio, the less diversified and more concentrated your portfolio becomes. The risk associated with each factor gets compounded.
Depends on how you measure diversification: # stocks versus # unique risks

Dave
rkhusky
Posts: 13510
Joined: Thu Aug 18, 2011 8:09 pm

Re: Why don't you factor tilt?

Post by rkhusky »

Random Walker wrote: Fri Sep 23, 2022 8:56 am
rkhusky wrote: Thu Sep 22, 2022 11:09 pm
km91 wrote: Thu Sep 22, 2022 3:17 pmThe larger the tilt away from the MKT premium and towards the other factors, the more diversified the portfolio should become.
Just the opposite is true. The more long-only factors you apply to your portfolio, the less diversified and more concentrated your portfolio becomes. The risk associated with each factor gets compounded.
Depends on how you measure diversification: # stocks versus # unique risks

Dave
Diversification in terms of different types of investments. For example, a portfolio of stocks, bonds, physical real estate, precious metals, commodities, micro-lending, and small businesses, is more diverse than just a stock + bond portfolio.

Likewise, a total market portfolio is more diverse than one with just stocks that have all the same factor characteristics.

Diversification reduces risk. Are you trying to say that a factor portfolio is less risky than TSM?
dbr
Posts: 42442
Joined: Sun Mar 04, 2007 9:50 am

Re: Why don't you factor tilt?

Post by dbr »

rkhusky wrote: Fri Sep 23, 2022 9:07 am

Diversification reduces risk. Are you trying to say that a factor portfolio is less risky than TSM?
Adding stocks to a portfolio of bonds increases diversification and increases risk -- or does it do either?
rkhusky
Posts: 13510
Joined: Thu Aug 18, 2011 8:09 pm

Re: Why don't you factor tilt?

Post by rkhusky »

dbr wrote: Fri Sep 23, 2022 9:17 am
rkhusky wrote: Fri Sep 23, 2022 9:07 am

Diversification reduces risk. Are you trying to say that a factor portfolio is less risky than TSM?
Adding stocks to a portfolio of bonds increases diversification and increases risk -- or does it do either?
Depends on how much you add. 10/90 has been less risky than 0/100.
dbr
Posts: 42442
Joined: Sun Mar 04, 2007 9:50 am

Re: Why don't you factor tilt?

Post by dbr »

rkhusky wrote: Fri Sep 23, 2022 9:37 am
dbr wrote: Fri Sep 23, 2022 9:17 am
rkhusky wrote: Fri Sep 23, 2022 9:07 am

Diversification reduces risk. Are you trying to say that a factor portfolio is less risky than TSM?
Adding stocks to a portfolio of bonds increases diversification and increases risk -- or does it do either?
Depends on how much you add. 10/90 is less risky than 0/100.
Has anyone ever presented a chart of risk vs diversification for stock/bond portfolios? The curve would be parameterized by asset allocation the same as an efficient frontier curve is. What would a curve of diversification as a function of asset allocation look like and what would a curve of risk as a function of asset allocation look like? What would be the definition of diversification that one could calculate and plot as a function of asset allocation?

This is intended to be a serious question.
dbr
Posts: 42442
Joined: Sun Mar 04, 2007 9:50 am

Re: Why don't you factor tilt?

Post by dbr »

rkhusky wrote: Fri Sep 23, 2022 9:37 am
dbr wrote: Fri Sep 23, 2022 9:17 am
rkhusky wrote: Fri Sep 23, 2022 9:07 am

Diversification reduces risk. Are you trying to say that a factor portfolio is less risky than TSM?
Adding stocks to a portfolio of bonds increases diversification and increases risk -- or does it do either?
Depends on how much you add. 10/90 has been less risky than 0/100.
I scanned some data for risk from Portfolio Visualizer and it turns out 20/80 is the most diversified stock/bond portfolio and 5/95 is the least risky portfolio by a tad. After 20/80 as diversification goes down risk goes up with more stocks.

The definition of diversification is the ratio of weighted average of the risks of the two components divided by the actual risk. Also it turns out the actual risk is very close to the root mean square of the risks consistent with stocks and bonds having zero correlation.
Random Walker
Posts: 5312
Joined: Fri Feb 23, 2007 8:21 pm

Re: Why don't you factor tilt?

Post by Random Walker »

dbr wrote: Fri Sep 23, 2022 12:59 pm
rkhusky wrote: Fri Sep 23, 2022 9:37 am
dbr wrote: Fri Sep 23, 2022 9:17 am
rkhusky wrote: Fri Sep 23, 2022 9:07 am

Diversification reduces risk. Are you trying to say that a factor portfolio is less risky than TSM?
Adding stocks to a portfolio of bonds increases diversification and increases risk -- or does it do either?
Depends on how much you add. 10/90 has been less risky than 0/100.
I scanned some data for risk from Portfolio Visualizer and it turns out 20/80 is the most diversified stock/bond portfolio and 5/95 is the least risky portfolio by a tad. After 20/80 as diversification goes down risk goes up with more stocks.

The definition of diversification is the ratio of weighted average of the risks of the two components divided by the actual risk. Also it turns out the actual risk is very close to the root mean square of the risks consistent with stocks and bonds having zero correlation.
The expected return of a portfolio is the weighted average of the portfolio component expected returns. The expected volatility of a portfolio is always less than the weighted volatilities of component volatilities because of correlations less than 1. Since volatility sucks at geometric returns, keeping expected return of a portfolio constant and decreasing portfolio volatility with low correlated components truly is potentially a bit of free lunch.

Dave
homebuyer6426
Posts: 966
Joined: Tue Feb 07, 2017 9:08 am

Re: Why don't you factor tilt?

Post by homebuyer6426 »

dbr wrote: Fri Sep 23, 2022 10:05 am
rkhusky wrote: Fri Sep 23, 2022 9:37 am
dbr wrote: Fri Sep 23, 2022 9:17 am
rkhusky wrote: Fri Sep 23, 2022 9:07 am

Diversification reduces risk. Are you trying to say that a factor portfolio is less risky than TSM?
Adding stocks to a portfolio of bonds increases diversification and increases risk -- or does it do either?
Depends on how much you add. 10/90 is less risky than 0/100.
Has anyone ever presented a chart of risk vs diversification for stock/bond portfolios? The curve would be parameterized by asset allocation the same as an efficient frontier curve is. What would a curve of diversification as a function of asset allocation look like and what would a curve of risk as a function of asset allocation look like? What would be the definition of diversification that one could calculate and plot as a function of asset allocation?

This is intended to be a serious question.
Is there an agreed upon standard for measuring diversification?
dbr
Posts: 42442
Joined: Sun Mar 04, 2007 9:50 am

Re: Why don't you factor tilt?

Post by dbr »

homebuyer6426 wrote: Fri Sep 23, 2022 2:30 pm
dbr wrote: Fri Sep 23, 2022 10:05 am
rkhusky wrote: Fri Sep 23, 2022 9:37 am
dbr wrote: Fri Sep 23, 2022 9:17 am
rkhusky wrote: Fri Sep 23, 2022 9:07 am

Diversification reduces risk. Are you trying to say that a factor portfolio is less risky than TSM?
Adding stocks to a portfolio of bonds increases diversification and increases risk -- or does it do either?
Depends on how much you add. 10/90 is less risky than 0/100.
Has anyone ever presented a chart of risk vs diversification for stock/bond portfolios? The curve would be parameterized by asset allocation the same as an efficient frontier curve is. What would a curve of diversification as a function of asset allocation look like and what would a curve of risk as a function of asset allocation look like? What would be the definition of diversification that one could calculate and plot as a function of asset allocation?

This is intended to be a serious question.
Is there an agreed upon standard for measuring diversification?
I don't know about agreement but I used the ratio of weighted mean of risk of components to actual risk. If the actual risk is less than the weighted mean of the components the measure of diversity is greater than one. I found in some actual data that the diversification of a portfolio of stocks and bonds runs from 1 for 1/100 to 1.4 at 20/80 back to 1 again at 100/0. So the most diversified portfolio is 20/80. The risk of a 100/0 portfolio was 3.96%, at 95/5 3.70% and then increases to 15.2%. That was a sample in Portfolio Visualizer using VBMFX and VTSMX.
Last edited by dbr on Fri Sep 23, 2022 3:29 pm, edited 1 time in total.
Random Walker
Posts: 5312
Joined: Fri Feb 23, 2007 8:21 pm

Re: Why don't you factor tilt?

Post by Random Walker »

homebuyer6426 wrote: Fri Sep 23, 2022 2:30 pm Is there an agreed upon standard for measuring diversification?
I don’t know, but I really like the concept of diversification ratio:
Mean weighted volatility portfolio components / portfolio volatility.

The diversification ratio thus incorporates the volatilities, correlations, and covariances of the portfolio components into a single number.

Dave
km91
Posts: 279
Joined: Wed Oct 13, 2021 12:32 pm

Re: Why don't you factor tilt?

Post by km91 »

dbr wrote: Fri Sep 23, 2022 3:17 pm I don't know about agreement but I used the ratio of weighted mean of risk of components to actual risk. If the actual risk is less than the weighted mean of the components the measure of diversity is greater than one. I found in some actual data that the diversification of a portfolio of stocks and bonds runs from 1 for 1/100 to 1.4 at 20/80 back to 1 again at 100/0. So the most diversified portfolio is 80/20. The risk of a 100/0 portfolio was 3.96%, at 95/5 3.70% and then increases to 15.2%. That was a sample in Portfolio Visualizer using VBMFX and VTSMX.
This is essentially risk parity where you hold stocks and bonds in a weighting where each component is contributing the same amount of risk to the portfolio. In fact, if you run the risk parity optimization on PV for VTSMX and VBMFX you will find that 20/80 is exactly the risk parity portfolio
User avatar
vineviz
Posts: 13325
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: Why don't you factor tilt?

Post by vineviz »

rkhusky wrote: Fri Sep 23, 2022 9:07 am Likewise, a total market portfolio is more diverse than one with just stocks that have all the same factor characteristics.

Diversification reduces risk. Are you trying to say that a factor portfolio is less risky than TSM?
Both of these statements are false.

A total stock market portfolio is certainly not necessarily "more diverse" than factor-diversified portfolio, and it's not hard to construct an equity portfolio that is more diversified than a market cap weighted equity portfolio.

And diversification does not necessarily reduce risk. Diversification is about balancing risk. De-risking reduces risk and can be accomplished with an increase, decrease, or no change at all in diversification.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
dcabler
Posts: 2602
Joined: Wed Feb 19, 2014 11:30 am
Location: Austin, TX

Re: Why don't you factor tilt?

Post by dcabler »

Random Walker wrote: Fri Sep 23, 2022 3:25 pm
homebuyer6426 wrote: Fri Sep 23, 2022 2:30 pm Is there an agreed upon standard for measuring diversification?
I don’t know, but I really like the concept of diversification ratio:
Mean weighted volatility portfolio components / portfolio volatility.

The diversification ratio thus incorporates the volatilities, correlations, and covariances of the portfolio components into a single number.

Dave
Same - Vineviz posted a sample spreadsheet a while back that does this nicely.

When you play around with it becomes kinda obvious that the holy grail is a portfolio composed of a multitude of assets, all having zero correlation and all having high returns. Good luck finding that. :D

The link to his spreadsheet is here: https://www.dropbox.com/s/n5vnejnvilfha ... .xlsx?dl=0

Cheers.
Last edited by dcabler on Fri Sep 23, 2022 6:16 pm, edited 5 times in total.
User avatar
HomerJ
Posts: 19773
Joined: Fri Jun 06, 2008 12:50 pm

Re: Why don't you factor tilt?

Post by HomerJ »

TightButAggressive wrote: Sun Sep 18, 2022 12:20 pm I have an interest in factor based investing and have read about it extensively. For those not familiar, the concepts have a long history with similar themes from Graham's Security Analysis all the way through to the Fama-French 5 factor model.

I have the following observations that have helped me reach a conclusion on factor based investing.

-There's always a backtest that shows outperformance based on some quantitative criteria. That does not mean the model will apply moving forward, no matter how logical it appears on face value.

-Not even the very smart folks can agree on exactly how to apply the concept. There's similar themes embedded in value investing, CAPM, MPT, Fama-French 3, FF5, Carhart 4, and many brokerage houses/market makers have their own funds, ETF's, and approaches. And many of these have later proven to have major flaws. That alone should signal that there is no foolproof approach. Most of all though, it signals a need for an active approach if (big if) you believe there's something to this concept as conventional wisdom will change.

-Most of the vehicles to apply factor based approaches have higher fees and/or expense ratios. I think most bogleheads will have the answer to whether or not higher cost is worth paying.

-I've looked at the returns of many different vehicles that are tied to the factors concept and they all underperform in various time horizons. Alternative weighting, smart beta, factor etf's, etc and I can't find a consistent set of returns that beats the market. People will of course find some time period with a specific approach that works, see point 1. But the variability in products and approaches is great. Extraordinary claims require extraordinary evidence.

Having said all of the above, I do think there are insights to be gained from factors. I also think there are a small subset of investors (3 sigma+) who are capable of leveraging quantitative and qualitative information to beat the market consistently. But the overwhelming conclusion for me on factors is that past performance doesn't prove future returns, there is no passive approach, there is a cost premium to taking this approach, and there is no single formula for how to apply it simply to generate market beating returns.

Ask yourself, are you a 3 sigma+ investor?
Excellent post. This is why I don't bother with factor tilting.
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
User avatar
Rowan Oak
Posts: 770
Joined: Mon May 09, 2016 2:11 pm
Location: Yoknapatawpha

Re: Why don't you factor tilt?

Post by Rowan Oak »

Jason Zweig had this to say today about why you shouldn’t factor tilt:
Along with value, momentum, low volatility and several other characteristics, quality is one of the factors that investment researchers say have generated higher performance than the market as a whole in the past.

It’s never certain, however, that investing in any particular one of these factors will provide a durably superior return in the future—especially given the perennial tendency of investors to jump in with both feet after a period of hot performance only to bail out when returns go cold.

If, on the other hand, you own all the factors all the time, then your returns will converge toward those of an index fund that holds the entire stock market—which you could own at a fraction of the cost of most factor funds.
https://www.wsj.com/articles/quality-st ... _permalink

note: This is a Wall Street Journal link which is usually behind a paywall, but I’ve had good luck viewing on mobile and on weekends.
“If you can get good at destroying your own wrong ideas, that is a great gift.” – Charlie Munger
User avatar
abuss368
Posts: 26709
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!
Contact:

Re: Why don't you factor tilt?

Post by abuss368 »

vineviz wrote: Fri Sep 23, 2022 3:39 pm
rkhusky wrote: Fri Sep 23, 2022 9:07 am Likewise, a total market portfolio is more diverse than one with just stocks that have all the same factor characteristics.

Diversification reduces risk. Are you trying to say that a factor portfolio is less risky than TSM?
Both of these statements are false.

A total stock market portfolio is certainly not necessarily "more diverse" than factor-diversified portfolio, and it's not hard to construct an equity portfolio that is more diversified than a market cap weighted equity portfolio.

And diversification does not necessarily reduce risk. Diversification is about balancing risk. De-risking reduces risk and can be accomplished with an increase, decrease, or no change at all in diversification.
Hi Vince -

Could you please provide an example of how diversification is different between total market index funds and factor funds?

Thanks.
Tony
John C. Bogle: “Simplicity is the master key to financial success."
User avatar
Taylor Larimore
Advisory Board
Posts: 31577
Joined: Tue Feb 27, 2007 8:09 pm
Location: Miami FL

Re: Why don't you factor tilt?

Post by Taylor Larimore »

Rowan Oak wrote: Fri Sep 23, 2022 4:38 pm Jason Zweig had this to say today about why you shouldn’t factor tilt:
Along with value, momentum, low volatility and several other characteristics, quality is one of the factors that investment researchers say have generated higher performance than the market as a whole in the past.

It’s never certain, however, that investing in any particular one of these factors will provide a durably superior return in the future—especially given the perennial tendency of investors to jump in with both feet after a period of hot performance only to bail out when returns go cold.

If, on the other hand, you own all the factors all the time, then your returns will converge toward those of an index fund that holds the entire stock market—which you could own at a fraction of the cost of most factor funds.
https://www.wsj.com/articles/quality-st ... _permalink

note: This is a Wall Street Journal link which is usually behind a paywall, but I’ve had good luck viewing on mobile and on weekends.
Rowan Oak:

Thank you for your post!

Jason Zweig is one of the most knowledgeable writers in the investment industry. He was our guest speaker at the second Boglehead Convention in 2001. Jason wrote this article that should help investors who are thinking about a "Factor Tilt":

I DON'T KNOW AND I DON'T CARE.

Best wishes.
Taylor

Jack Bogle's Words of Wisdom: “It’s very difficult for any particular segment of the stock market to sustain superior performance. The watch word for our financial markets is ‘reversion to the mean.’”
"Simplicity is the master key to financial success." -- Jack Bogle
Nathan Drake
Posts: 3808
Joined: Mon Apr 11, 2011 12:28 am

Re: Why don't you factor tilt?

Post by Nathan Drake »

If, on the other hand, you own all the factors all the time, then your returns will converge toward those of an index fund that holds the entire stock market—which you could own at a fraction of the cost of most factor funds.
I believe this is an incorrect interpretation of factor investing. If you own the additional risk factors, you will get the market factor + whatever these additional risk premia generate based on how much loading of each factor they have. They do not "converge" to an index fund by owning all of them.
20% VOO | 20% VXUS | 20% AVUV | 20% AVDV | 20% AVES
User avatar
enad
Posts: 219
Joined: Fri Aug 12, 2022 2:50 pm

Re: Why don't you factor tilt?

Post by enad »

HomerJ wrote: Fri Sep 23, 2022 4:22 pm
TightButAggressive wrote: Sun Sep 18, 2022 12:20 pm I have an interest in factor based investing and have read about it extensively. For those not familiar, the concepts have a long history with similar themes from Graham's Security Analysis all the way through to the Fama-French 5 factor model.

I have the following observations that have helped me reach a conclusion on factor based investing.

-There's always a backtest that shows outperformance based on some quantitative criteria. That does not mean the model will apply moving forward, no matter how logical it appears on face value.

-Not even the very smart folks can agree on exactly how to apply the concept. There's similar themes embedded in value investing, CAPM, MPT, Fama-French 3, FF5, Carhart 4, and many brokerage houses/market makers have their own funds, ETF's, and approaches. And many of these have later proven to have major flaws. That alone should signal that there is no foolproof approach. Most of all though, it signals a need for an active approach if (big if) you believe there's something to this concept as conventional wisdom will change.

-Most of the vehicles to apply factor based approaches have higher fees and/or expense ratios. I think most bogleheads will have the answer to whether or not higher cost is worth paying.

-I've looked at the returns of many different vehicles that are tied to the factors concept and they all underperform in various time horizons. Alternative weighting, smart beta, factor etf's, etc and I can't find a consistent set of returns that beats the market. People will of course find some time period with a specific approach that works, see point 1. But the variability in products and approaches is great. Extraordinary claims require extraordinary evidence.

Having said all of the above, I do think there are insights to be gained from factors. I also think there are a small subset of investors (3 sigma+) who are capable of leveraging quantitative and qualitative information to beat the market consistently. But the overwhelming conclusion for me on factors is that past performance doesn't prove future returns, there is no passive approach, there is a cost premium to taking this approach, and there is no single formula for how to apply it simply to generate market beating returns.

Ask yourself, are you a 3 sigma+ investor?
Excellent post. This is why I don't bother with factor tilting.
Jack Bogle also said: Financial history is important, and studying historical rates of returns provides useful perspectives. But beware of concluding too much from the past returns in the financial markets. Beware of past returns for periods that seem long enough but are not.

Portfolio Visualizer is often used to prove a point. Sometimes I click on the link to see the point someone is trying to make and the last time the person used 2000-2019 to prove their point. I changed the start/end dates by 1 year and the point could not be made.
Watch me pull a rabbit out of my hat
User avatar
vineviz
Posts: 13325
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: Why don't you factor tilt?

Post by vineviz »

enad wrote: Fri Sep 23, 2022 6:53 pm
Jack Bogle also said: Financial history is important, and studying historical rates of returns provides useful perspectives. But beware of concluding too much from the past returns in the financial markets. Beware of past returns for periods that seem long enough but are not.
Unless you need those past returns to justify a US-only equity tilt, in which case go ahead and embrace them.

More to the point, none of the evidence used to build our modern asset pricing models could aptly be characterized as simply "studying historical rates of returns" as Bogle understood it.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
rkhusky
Posts: 13510
Joined: Thu Aug 18, 2011 8:09 pm

Re: Why don't you factor tilt?

Post by rkhusky »

vineviz wrote: Fri Sep 23, 2022 3:39 pm
A total stock market portfolio is certainly not necessarily "more diverse" than factor-diversified portfolio, and it's not hard to construct an equity portfolio that is more diversified than a market cap weighted equity portfolio.
And is your constructed equity portfolio a standard factor portfolio with all the stocks that have the same factor characteristics? For example, is a SCV fund more diverse than a TSM fund?
User avatar
abuss368
Posts: 26709
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!
Contact:

Re: Why don't you factor tilt?

Post by abuss368 »

enad wrote: Fri Sep 23, 2022 6:53 pm
Jack Bogle also said: Financial history is important, and studying historical rates of returns provides useful perspectives. But beware of concluding too much from the past returns in the financial markets. Beware of past returns for periods that seem long enough but are not.

Portfolio Visualizer is often used to prove a point. Sometimes I click on the link to see the point someone is trying to make and the last time the person used 2000-2019 to prove their point. I changed the start/end dates by 1 year and the point could not be made.
[/quote]

Hi enad -

Unfortunately that is one disadvantage of looking at historical data. Change the input variables by a small amount and the results can be significantly different.

Best.
Tony
John C. Bogle: “Simplicity is the master key to financial success."
User avatar
abuss368
Posts: 26709
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!
Contact:

Re: Why don't you factor tilt?

Post by abuss368 »

vineviz wrote: Fri Sep 23, 2022 7:03 pm
enad wrote: Fri Sep 23, 2022 6:53 pm
Jack Bogle also said: Financial history is important, and studying historical rates of returns provides useful perspectives. But beware of concluding too much from the past returns in the financial markets. Beware of past returns for periods that seem long enough but are not.
Unless you need those past returns to justify a US-only equity tilt, in which case go ahead and embrace them.

More to the point, none of the evidence used to build our modern asset pricing models could aptly be characterized as simply "studying historical rates of returns" as Bogle understood it.
Vince -

You have me very curious up thread regarding your diversification note. Could you please provide an example of how diversification is different between total market index funds and factor funds?

Thanks.
Tony
John C. Bogle: “Simplicity is the master key to financial success."
User avatar
abuss368
Posts: 26709
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!
Contact:

Re: Why don't you factor tilt?

Post by abuss368 »

Taylor Larimore wrote: Thu Sep 22, 2022 7:21 pm Bogleheads:

For what it's worth, I looked-up the longest Morningstar return (3 years) for Vanguard U.S. Multifactor ETF (VMMF). It was 8.49%.

Meanwhile the 3 year return for Vanguard U.S. Total Market Index ETF (VTI) was 9.34%. Another reason I'm happy I don't "factor tilt."

What Experts Say About Total Market Index Funds

Best wishes.
Taylor
Jack Bogle's Words of Wisdom: "One of the seemingly indestructible myths of investing is that stocks with small market capitalizations outpace stocks with large market capitalizations over time."
Hi Taylor -

Thank you for sharing that performance data. While three years is a very short time period, when using these strategies, investors have to be willing to stay the course over a very long period of time to hopefully have that strategy pay off.

Hope you are well.

Tony
John C. Bogle: “Simplicity is the master key to financial success."
shriram
Posts: 38
Joined: Sat Oct 07, 2017 12:10 am

Re: Why don't you factor tilt?

Post by shriram »

Unfortunately that is one disadvantage of looking at historical data. Change the input variables by a small amount and the results can be significantly different.
Make sense! Btw I backtested returns between Vanguard Total Stock Market index fund vs DFA Small Cap Value fund since 01/1994 - 08/2022. This is longest period for which data is available. If you look at the graph, the DFA fund has almost always performed better than total stock market index fund (except for 3 years between 1998 - 2001).

Here is the backtest result: https://www.portfoliovisualizer.com/bac ... ion2_2=100

Looking back, for a long term investor (with 20+ years time horizon) - it would certainly have made sense to invest portion of their investment in SCV during 01/1994. As long as they did not capitulate before/during dot com crash (1998-2001), they would have come ahead in terms of returns by investing SCV against Total Stock Market index.

So to me, the main pre-requisite for investing in SCV is
(1) to have long term time horizon for the investment
(2) patience to stick with the fund even during downturn.

But aren't these two aspects main part of the Bogleheads philosophy ?
km91
Posts: 279
Joined: Wed Oct 13, 2021 12:32 pm

Re: Why don't you factor tilt?

Post by km91 »

Nathan Drake wrote: Fri Sep 23, 2022 5:24 pm
If, on the other hand, you own all the factors all the time, then your returns will converge toward those of an index fund that holds the entire stock market—which you could own at a fraction of the cost of most factor funds.
I believe this is an incorrect interpretation of factor investing. If you own the additional risk factors, you will get the market factor + whatever these additional risk premia generate based on how much loading of each factor they have. They do not "converge" to an index fund by owning all of them.
TSM will by definition hold all stocks with the characteristics of the factors by cap weight. This does not give an investor significant exposure to the factors. What we want is to hold stocks with factor characteristics overweight relative to TSM. The point is to diverge from a market cap weighted portfolio and to introduce tracking error relative to TSM. And hope that we are compensated for the tracking error we are taking
User avatar
vineviz
Posts: 13325
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: Why don't you factor tilt?

Post by vineviz »

abuss368 wrote: Fri Sep 23, 2022 7:45 pm
You have me very curious up thread regarding your diversification note. Could you please provide an example of how diversification is different between total market index funds and factor funds?
It's well-trodden ground, but I'll briefly recap.

Diversification is a process of balancing the risk exposures of a portfolio across multiple sources.

A total market index fund has just one risk factor exposure, which is the market factor.

A more diversified fund might have exposure to the market factor AND other sources of risk & return, like value or profitability or size. To the extent that those factors are independent (i.e. less than perfectly correlated with each other) there's a chance that when one factor is performing badly another factor can help offset it.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
vineviz
Posts: 13325
Joined: Tue May 15, 2018 1:55 pm
Location: Baltimore, MD

Re: Why don't you factor tilt?

Post by vineviz »

rkhusky wrote: Fri Sep 23, 2022 7:36 pm
vineviz wrote: Fri Sep 23, 2022 3:39 pm
A total stock market portfolio is certainly not necessarily "more diverse" than factor-diversified portfolio, and it's not hard to construct an equity portfolio that is more diversified than a market cap weighted equity portfolio.
And is your constructed equity portfolio a standard factor portfolio with all the stocks that have the same factor characteristics? For example, is a SCV fund more diverse than a TSM fund?
It could be. Obviously depends on the funds you're comparing.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
Post Reply